The discipline of multiple attenuation continues to make significant steps forward in the ultimate goal of eliminating multiples from various types of seismic data. As a retrospective on this, we refer back to the last TLE special section on multiple attenuation (January 1999) and compare it to the papers in this issue and the recent literature. The 1999 issue featured significant progress in the attenuation of multiples using wave-equation-based methods, which are popularly referred to as surface-related multiple elimination or SRME. A significant portion of that section was populated with papers from a multiple attenuation workshop held at SEG's 1997 Annual Meeting in Dallas. At that workshop, there was a growing acceptance that SRME was not only a viable method, but that it could attenuate multiples that previous methods could not. This was, of course, the promise of theory borne out in practice; however, one important fly in the ointment was that all applications were limited to 2D. While hopeful, the tone at the end of the workshop was: What about 3D SRME?